This is part of my M.Tech. Thesis work. Main highlights are:

• Achieved up to 81.6% accuracy in Diabetes Prediction on Pima Indians Diabetes Database with Random Forest classifier
• Applied and analysed accuracies of “K-Nearest Neighbors, Support Vector Machine, Decision Tree and Random Forest” classification algorithms for diabetes prediction
• Achieved up to 7.04% improvement in the accuracy of Decision Tree classification algorithm for Diabetes Prediction
• Predicted missing values present in the dataset using a set of “Linear Regression, Support Vector Regression, Decision Tree and Random Forest” regression algorithms
• Performed Dataset Balancing using SMOTE algorithm and then Feature Scaling